ALTA's 'rating' system

Last week I blogged about USTA’s rating system which, while it has its flaws, is about as good as it gets.  It’s certainly superior to having no rating system at all, and the relativist one that ALTA employs.

ALTA’s rating system, if you could call it one, uses its own historical team-based data to place teams at different levels.  Instead of rating individuals, all the players on a given team are assigned the same level which is then adjusted based upon that team’s performance at the end of the season.  In other words, a team comprised of a mix of players with USTA abilities ranging from 2.5 to 4.5 levels could be placed at the B-5 level, where there would likely be mismatches, week in and week out, and virtually every line.  This is more common than you think - especially in mixed doubles - because generally ALTA teams are organized in neighborhoods, and not all communities have an even mix of enough players of similar ability to form a team.

A group of highly rated USTA players who have never played ALTA – and hence, don’t have ALTA ratings – will likely be placed too low initially and dominate until their ALTA level matches their ability.  This will affect multiple levels for several seasons until the team finds its equilibrium point within ALTA’s system.  Perhaps your team has been improving, even hired a coach in the offseason, and is finally at a point to compete for a bag tag when, all of a sudden, a new team is placed within your division that dominates all comers.  Oh well, sucks to be you.  Although ALTA’s relative placement system eventually puts teams where they belong, individual teams will suffer along the way.  Although I’m not sure how it could be implemented, it would be nice if ALTA would utilize USTA (or even other league) ratings to place teams initially, so that a brand new team doesn’t end up winning plates at the (e.g.) C-8, C-1, B-4 & A-8 levels before they ‘re finally placed at their high A level.  I have no idea if this has ever happened.

The only leveling that occurs in ALTA’s ‘system’ is when players want to change teams.  Their last team level is assigned a point value which, when added to the point values of the top 9 players on a team, is used to re-level their new team.  Of course, the fact that the specific A-7 player that now wants to join a B-3 team hardly ever played and/or wasn’t very competitive at that A level doesn’t enter into ALTA’s decision; hence, the B-3 team could be bumped to B-2 without really improving their team’s roster.  The reverse scenario almost never helps a team to move down a level because, unless they lose all but 9 players on their previous team, the top 10 players have the same assignment that the previous season’s performance had earned them.  Therefore, the ‘system’ is geared to move teams up, which I guess is the objective:  players only get better by playing better players.  But for players whose abilities are declining with age, it can be hard sledding for them and their teams.

I wrote previously about ALTA’s need for transparency, something which would allow everyone to better see the disparities inherent in the league’s rating system and would also expose the neighborhoods that exploit it to their advantage to win titles, so I’ll not regurgitate this travesty further.  But I will suggest a serious look at revamping the system.

Rather than just criticize what doesn’t work ideally, I’ll offer a suggestion for improvement.  ALTA should start to rate individual players, not just teams.  If you’ve ever been a captain of a team that has made the playoffs, you know that ALTA has a system to keep a team from sandbagging once they’re in.  Every player is assigned a value for the line that they played each week during the season which takes into account how they did.  For example, a line 1 player is given a 1.0 for a week in which he and his partner won or a 1.5 if they lost, and so forth for lines 2 through 5.  That player’s cumulative total is then added to their partner’s in the captain’s proposed lineup such that the lowest numbered pairing has to be placed highest, the second lowest numbered pairing is placed at line 2, etc.  Why not keep and use this historical seasonal data to assist in placing teams, and especially when players change teams?  Also, keep the data so that players can’t sit out a year and reenter with no rating, my chief complaint with the large communities that abuse the ‘system’.

I think that ALTA can improve/establish a rating system that would work better for everyone, but I don’t know if they’re interested or are willing to put forth the effort or spend the money needed to do it.  If not, I believe that it will continue to be a source of dissatisfaction in what is otherwise a premier league.